The team builds the core software behind GPU virtualization for AI and machine learning workloads, and you would work on low-level C++ code that raises GPU utilization and supports a growing cloud infrastructure platform.
Responsibilities
Build and upkeep a C++ library for virtualization.
Study approaches to GPU oversubscription, checkpointing, and high-performance system design.
Add support for newer hardware architectures.
Track down bugs that span multiple layers of the stack when they appear in production.
Find the causes of slow performance in customer workloads and fix them.
Requirements
You have at least 3 years of C++ systems work and strong C++ skills. Rust experience also counts.
You have run low-level software in production, such as operating systems, kernels, memory management, or drivers.
You have improved network interface or C++ performance and traced slowdowns through several system layers.
You have shipped production software, ideally at an early-stage company.
You have solid computer science fundamentals and a relevant degree.
Experience with compilers, networking protocols, Kubernetes, infrastructure, distributed systems, storage, or data ingestion is valuable.
Conditions
Base salary is $250,000 to $325,000 per year.
Visa sponsorship is available.
The role is on-site in San Francisco, California, United States.